MGI Tech launched Genoria AI, a new AI-focused subsidiary designed to develop and commercialize “AI lab agent” tools that can plan and execute biomedical experiments autonomously. The company says the lab-agent approach combines large language models with laboratory automation platforms, with an emphasis on closed-loop execution from experimental goals to instrument-executable protocols. MGI’s chief AI officer, Meng Yang, leads Genoria AI and described ProtoPilot as the underlying model for converting scientific goals into automated workflows, which then learn from outcomes for iterative refinement. The company demonstrated proof-of-concept on multiple intent-to-code tasks and wet-lab workflows in a preprint. For biotech and lab operations, the step is a concrete move toward software-driven research execution, with potential implications for throughput in discovery pipelines and standardization across experimental setups.